Progress Note Understanding - Assessment and Plan Reasoning: Overview of the 2022 N2C2 Track 3 shared task

Area of Science:

• Medical Informatics
• Natural Language Processing
• Clinical Decision Support

Background:

• Electronic Health Records (EHRs) contain extensive patient data, but 'note bloat' can obscure critical information.
• Natural Language Processing (NLP) is increasingly applied to EHRs, primarily for information extraction.
• Few NLP applications currently support downstream diagnostic decision-making.

Purpose of the Study:

• Introduce the 2022 National NLP Clinical Challenge (N2C2) Track 3: Progress Note Understanding - Assessment and Plan Reasoning task.
• Develop and evaluate NLP systems to predict causal relations between patient status (Assessment) and diagnoses/treatments (Plan) in progress notes.
• Enable automated prioritization of diagnoses for diagnostic decision support.

Main Methods:

• Focused on the Assessment and Plan subsections of daily progress notes.
• Task required NLP systems to identify causal links between patient status and specific diagnostic/treatment components.
• Utilized data from the N2C2 2022 Track 3 challenge.

Main Results:

• Presents the outcomes of the N2C2 2022 Track 3 challenge.
• Details system performance based on the defined evaluation metrics.
• Provides insights into the capabilities of NLP for progress note analysis.

Conclusions:

• The Assessment and Plan Reasoning task is a crucial step towards advanced diagnostic decision support.
• Automating the identification and prioritization of diagnoses can enhance clinical workflow.
• Further development in NLP for EHRs can help mitigate challenges posed by large clinical documents.